RSS usability guidelines

i started to consume news and weblogs via RSS only some months ago but i’ve been addicted ever since. i use bloglines.com as feed aggregator, as i use multiple computers and therefor a website as RSS client seems more suitable than a program. (by the way, bloglines clearly has become the website i use most, basically it’s open in the background all the time.) see the public view of the feeds i’m subscribed to on bloglines.

RSS certainly is the next big thing (or already the current one), i really believe it will play a major role in publishing in the future. the days of email newsletters certainly are about to be over.

however, as RSS is young and not all publishers seem to consume the feeds they produce themselves, there are still a lot of teething problems. i have taken notes of common feed publishing errors that caught my attention, so if you consider adding a RSS feed to your website or weblog you might want to use my list as a RSS usability guideline. in case you are the author of one of the feeds quoted (which are all great and recommendable – they wouldn’t be on my blogroll otherwise!) kindly understand that as constructive feedback, and feel free to drop a comment below.

update: see end of article.

list of common feed publishing errors

1) disguised links
i don’t know if they do it on purpose (to route traffic to their sites?), due to technical restrictions of the software they use or by accident: many feed authors disguise links in their texts. that can make them very much confusing and unreadable – to understand what the author wants to say, i often have to go to their website and read the entry there. take for example this blog entry and look at the first link. it was not underlined/clickable in my feed aggregator, you can imagine how this impairs the reading experience. others on my blogroll committing the same error: mask.at, thomas n. burg’s randgaenge.net, voice in the desert and web log tools collection.

2) too little text
some RSS feeds simply display too little flesh, not enough to really know what the article is about and if it’s worth reading. examples: orf futurezone, jakob nielson’s useit.com (interestingly the usability expert has been having a typo in his own name for many weeks now – in the feed’s very title!), martin röll’s e-business weblog, randgaenge.net and telepolis news. especially this point is highly subjective, of course!

3) quotations not clearly marked as such
it has become generally accepted to clearly mark literal excerpts from external sources, often they are indented (i have to admit, i still have to create an appropriate CSS quotation class on helge.at..). yet often these excerpts aren’t marked at all in the RSS versions, so as a reader i can’t distinguish the publisher’s words from the quoted text. (sometimes when the punctuation is missing the text becomes even unreadable, as quotation and the rest of the article form new, senseless sentences.) examples from my list: louis rosenfeld’s “bloug”, googleguy says, highly overrated and soulsoup.

4) updates not marked as such
RSS-clients display articles in a simlar way email clients display mails: i can keep track of what articles i’ve read already as unread articles are marked. the problem is that bloglines (probably just as any other reader too, but i’m not sure) treats changed articles just as new ones. i hope RSS-clients will be able to deal with this problem and highlight changed parts somehow. until this is the case i would advise feed publishers to highlight the changes they make themselves (e.g. by adding “update: typo fixed” to the end of the article). an extreme example of this elsewhere minor problem is golem.de: they add “(NN comments)” to the end of each feed, where NN obviously increases all the time, which causes already read articles to bump all the time. the only case of a RSS publishing error that actually makes me consider unsubscribing!

5) too many articles per day
some feeds i’m subscribed to have a far too large output of articles which makes it hard to keep track. e.g. bbc-news africa has published 250 articles in the last 3 weeks, most of them about african football, which i’m absolutely not interested in. it would be advisable to split such a feed in subfeeds such as “african sports” or “african politics”. allafrica.com burkina faso news in french don’t just publish too many articles (many of just very local importance), they also usually publish large numbers articles all at once (every second day or so), which gives me a huge list even though i check bloglines several times a day.

6) links to access-restricted content not marked as such
the specialists’ community webmasterworld.com has some public forums and one forum that is restricted to paying subscribers – yet articles aren’t distinguishable in the RSS version until i follow a link just to maybe find out that i don’t have access. it would be better (and easy) to display the forum name the thread resides in, along with a note “paying subscribers only”, or so.

7) nested feeds
i’m both subscribed to the register – mobile and the register – wlan. unfortunately the wlan feed is a subset of the mobile feed, so i get to read all wlan articles twice. ok, i could unsusbcribe from the wlan feed, but i only want to quickly scan all those mobile articles while the wlan articles are of greater interest. a minor issue, i admit (but avoidable). a workaround would be if RSS clients would offer the ability to highlight words defined by me within a feed (“wlan” in the mobile feed, in this case). meant as a suggestion to the bloglines folks (post a comment in case this has reached you! ;-)

8) confusion between authors and categories
sometimes a weblog software’s author function is abused to simulate categories (as also in my installation of greymatter) – or there might be other reasons that some feeds display strange author and/or category information. although i remember having seen this somewhere else too, the only example i can find is the knallgrau company weblog which displays “by authorname on authorname” (only in RSS of course, not on their website) instead of “by authorname on category”. again, a minor issue.

there are probably more RSS feed errors that are common (and certainly RSS-client-specific ones), those are just the ones i’ve stumbled accross so far.

useful links:
» RSS tutorial for content publishers and webmasters
» plugging the RSS usability hole (a usability error my website commits as well, i have to admit)

update july 22: thomas n. burg has – coincidentially, almost immediately after this post – fixed one of the two flaws i attributed to his blog, text-length has been extended to 1000 characters.
update july 28: christian klaß, founder of golem.de, has sent an email: “bitte nochmal checken, seit gestern müsste alles korrekt sein.”